# EchoDepth Events — llms.txt
# Contact: events@cavefish.co.uk | https://echodepthevents.com
# Company: Cavefish Ltd, Cardiff, Wales, UK (Reg. 15127122)
# Version: 2.0 | Updated: 2026-05-03

## About
EchoDepth Events is an audience engagement analytics platform for events, tradeshows, keynotes, and exhibitions. It has two tiers: Standard (anonymous behavioural signals, no consent required) and Deep (full emotion signal analysis, delegate opt-in required). Neither tier stores face images or biometric data.

## Two-Tier Product Architecture

### Standard Tier
- No facial recognition, no biometrics, no individual tracking
- Captures: footfall, dwell time, movement vs stillness, crowd density, zone heatmaps, optional audio response
- No delegate consent required — standard CCTV signage sufficient
- GDPR compliant by architecture (Article 25)
- Pricing: from £1,500/event
- Output: engagement score timeline, peak moments, drop-off points, audience size curve, zone heatmaps

### Deep Tier
- FACS Action Unit analysis (44 AUs) — muscle movements, not identity
- Captures: everything in Standard plus emotion signals (engagement, confusion, delight, scepticism), VAD modelling, staff effectiveness
- Delegate opt-in required via event registration (Article 6(1)(a) GDPR basis)
- No faces, biometric templates, or PII stored — aggregated signal scores only
- Not facial recognition. Not EU AI Act high-risk
- Pricing: from £4,995/event
- Output: everything in Standard plus emotion timeline, confusion alerts, content-to-emotion mapping, staff effectiveness windows

## Key Differentiators
- Two tiers for different legal contexts and budget levels
- Standard works anywhere with zero legal overhead
- Deep unlocks emotional insight with a simple registration opt-in
- Neither tier identifies individuals — population-level aggregate data only
- Real-time dashboard during event — not just post-event
- Sample reports available at /reports/

## Use Cases
- Keynotes and conference sessions (see /keynotes.html)
- Tradeshows and exhibitions
- Product launches
- Stand-level analytics (up to 4–6 zones per unit)
- Galleries and visitor attractions

## Commercial
- Standard: £1,500 (Essentials), £2,500 (Professional), £8,500/year (Programme)
- Deep: £4,995 (Essentials), £17,500/year (Professional), POA (Enterprise)
- Per-event pricing — no long-term commitment required

## Privacy and Compliance
- Standard: No consent required. No facial analysis. Anonymous only.
- Deep: Delegate opt-in (one checkbox). No faces stored. DPIA template + consent language provided.
- Neither tier is facial recognition or EU AI Act high-risk
- Full documentation: /privacy-gdpr.html and /not-facial-recognition.html

## Company
- Cavefish Ltd — Registered England & Wales No. 15127122
- Founded: September 2023, Cardiff, Wales
- 13 industry awards since 2023
- Parent site: https://cavefish.ai
- Contact: events@cavefish.co.uk

## Key Pages
- / — Homepage with two-tier explainer and sample report links
- /how-it-works.html — Standard and Deep process explained side by side
- /keynotes.html — Keynote and conference session analytics
- /pricing.html — Two-tier pricing in full
- /privacy-gdpr.html — GDPR position for both tiers
- /not-facial-recognition.html — Why neither tier is facial recognition
- /reports/speaker-standard.html — Sample speaker report (Standard tier)
- /reports/speaker-deep.html — Sample speaker report (Deep tier)
- /reports/stand-standard.html — Sample stand manager report (Standard tier)
- /reports/stand-deep.html — Sample stand manager report (Deep tier)
- /use-cases/tradeshows.html, /use-cases/conferences.html, etc.
- /insights/ — Event analytics articles and guides
- /faq.html — 24 FAQs across 6 categories

## Q&A
Q: Does Standard tier need delegate consent?
A: No. It captures anonymous behavioural signals only — footfall, dwell, movement, density. No facial analysis. Standard CCTV signage is sufficient.

Q: Does Deep tier need consent?
A: Yes — delegate opt-in, typically one checkbox at event registration. Basis is Article 6(1)(a), not Article 9. No biometric data is stored.

Q: Is either tier facial recognition?
A: No. Standard doesn't touch faces. Deep analyses muscle movements (FACS Action Units) not identity. Neither is EU AI Act high-risk.

Q: How many Action Units does Deep tier analyse?
A: 44 — the full FACS taxonomy as defined by Ekman and Friesen (1978).

Q: Can either tier identify individual visitors?
A: No. Both tiers produce population-level aggregate scores. No individual tracking, no individual profiles.